Growing Up in Ireland Conference

The Value of Child Cohort Studies –

Historical and International Perspectives

Professor John Bynner, Centre for Longitudinal Studies, Institute of Education, London

7th December 2009, Dublin

Contents

1. Why longitudinal research

2. Longitudinal studies: time and place

3. The studies

4. Context of social change

5. Changing life course process

6. Trajectories of disadvantage and policy challenge

Value of longitudinal data

Predicting consequences of early experience and circumstances – e.g. childhood disadvantage

Explaining outcomes – e.g. Not in Education, Employment or Training (NEET)

Estimating returns – e.g. to qualifications

Identifying factors that override predictions – e.g. “escape from disadvantage”

Life course dynamics – e.g. interactions between literacy proficiency, ICT competence, employment, family, health

Bynner, J. & Joshi, H. (2007) ‘Building the Evidence Base from Longitudinal Data’. Innovation, 20, 159-179

Bynner et al 2009 Use of Longitudinal Research in the evaluation of the Scottish Government’s Strategic Outcomes http://www.scotland.gov.uk/Publications/2008/11/25160140/0

e.g.2000 e.g. 2010

Types of longitudinal cohort study

Birth cohort studies: 1946, 1958, 1970, 1992 Avon Longitudinal Survey of Parents and Children (ALSPAC), 2000 Millennium Cohort Study (MCS)

2005 Growing Up in Scotland (GUS), 2008 Growing Up In Ireland (GUI), 2010 French Birth cohort study (ELFE), 2012 German National Educational Panel Study (NEPS contains one birth cohort out of multiple age cohorts), 2009 US National Children’s Study (NCS)

Age cohort studies: 2004 Longitudinal Study of Young People in England (LSYPE), 2005 Growing Up in Scotland (GUS), 2008 Growing Up in Ireland (GUI), 2012 German National Educational Panel Study (NEPS)

Sequential cohort studies: UK Youth Cohort Study (YCS), Scottish School Leavers Survey (SLS), 2012 German National Educational Panel Study (NEPS)

Martin, J. Bynner, J. Kalton, G. Boyle, P. Goldstein, H. Gayle, V. Parsons, S. Piesse (2006), A. Review of Panel and Cohort Studies . Bynner, J. Wadsworth, M Goldstein, H. Maughan, B. Purdon, S. Michael, R. (2007), Scientific Case for a New Birth Cohort Study. Bynner, J. Wadsworth, M Goldstein, H. Maughan, B. Lessof, C. Michael, R. (2009) Options for the design of the 2012 birth cohort study. www.Longviewuk.com /pages/reportsnew.shtml

British Birth Cohort Studies

54 Year NSHD

NSHD

05

1015202530354045505560

1940 1950 1960 1970 1980 1990 2000 2010

British Birth Cohort StudiesNational Child Development Study (NCDS)

NCDS

7 Year NCDS

11 Year NCDS

23 Year NCDS

33 Year NCDS

42 Year NCDS

50 Year NCDS

05

1015202530354045505560

1940 1950 1960 1970 1980 1990 2000 2010

37 Year NCDS

Child

Data

16 years NCDS

British Birth Cohort Studies1970 British Cohort Study (BCS70)

22 Months BCS7042 Months BCS70

BCS70

10 Years BCS70

21 Years BCS70S (sub sample)

30 Years BCS70

5 Years BCS70

26 Years BCS70

38 Years BCS70

05

1015202530354045505560

1940 1950 1960 1970 1980 1990 2000 2010

16 Year BCS70

British Birth Cohort StudiesMillennium Cohort Study MCS)

051015202530354045505560

1940 1950 1960 1970 1980 1990 2000 2010

9 months 3 years

7 years5 years

11 years

MCS

Age

The British Birth Cohort studies

NCDS 1958 (n = 17,000)

BCS70 1970 (n = 17,000)

MCS 2000 (n = 19,000)

NSHD 1946 (n = 5,000)

051015202530354045505560

1940 1950 1960 1970 1980 1990 2000 2010

Child

Data

Age

Year

Child Data

ALSPAC 1992 (n = 15,000)

.

NCS 2012

(n= 30,000)

NCS= New cohort study

Theoretical perspective

Distal and Proximal ecological factors in child development (Bronfenbrenner)

Interactional contexts – e.g. the family

Structural factors - e.g. social class, locality

Life course trajectories (Elder, Heinz) shaped by :

Human agency – development of the individual

Linked-lives – social relations

Timing – age, period, cohort

Location in time and space – history and culture

Housing space – people per room

%

Birth cohort

Parents/carers with degrees

Birth cohort

%

Experienced temporary suspension from school by family social class

Ferri, E. , Bynner, J. and Wadsworth . M. (2003) Changing Britain Changing Lives . IoE press

%

Birth cohort

unskilled

professional

Modeling the routes to adult statuses

Early resource accumulation Transition resources Adult statuses

Family background Early influences Leaving choicesEmployment/Family

(Birth) (Age 10/11) (Age 16+) Age 26

Rural vs urban area

Structural equation model of education pathways to unemployment

Exogenous (x) Endogenous (y) Outcome (y)

Before 10 10-11 16+ 21/23

Mother left

education

Father left education

Family social class

Maths

Reading

Mother’s interest

Father’s interest

Overcrowding

Exams score

Literacy score

Numeracy score

Age left education

Number of jobs

Malaise

No work-based training

Unemployment 16-21 / 16-23

1958 cohort boys: impact of family background

Before 10 11

Maths

Reading

Mother’s interest in education

Father’s interest in education

Overcrowding

Rented housing

Social class

.11

.19

.35

.43

-.19

-.50

1970 cohort boys: impact of family background

Before 10 10

Age mother left education

Age father left education

Maths

Reading

Mother’s interest in education

Father’s interest in education

Overcrowding

Rented housing

Social class

.13

.08

-.19

-.15

.13

-.14

.11

.08

.10

.11

1958 cohort boys: influences on school leaving exam scores

11 16+

Maths

Reading

Overcrowding

Exams score

.46

.22

-.06

1970 cohort boys: influences on school leaving exam scores

10 16+

Maths

Reading

Rented housing

Father’s interest

Overcrowding

Exams score

.30

.21

.10

-.12

-.15

1958 cohort boys: influences on unemployment

16+ 16-23

Exams score

Numeracy score

Number of jobs

Unemployment 16-23

-.11

-.12

.17

1970 cohort boys: influences on unemployment

16+ 16-21

Exams score

Literacy score

Numeracy score

Number of jobs

Malaise

Social class

Unemployment 16-21

-.16

-.11

-.10

.09

.10

-.11

Trajectory of disadvantage Age Disadvantage Education outcome Intervention

0-13

Unskilled family, Overcrowded, rented home

Free school meals

Workless family

Lack of parental interest in child’s education

Poor pre school cognitive skills

Literacy and numeracy development slow

Falling behind at school

Every Child Matters

Sure Start

14-22

Casual unskilled work

NEET

Poor literacy and numeracy

No qualifications

Leave school early

Educational maintenance allowances

Connexions

Extend education inside and outside institutions

School age 18 graduation

25+ Out of labour market/ early parenthood (Women)

Unemployment/ delayed partnership (Men)

No employer-based training

Prospects poor

Skills for Life

Provision matched to situation and needs

Embedded curriculum,

0-13

Disadvantaged circumstances

Cycle repeated Break the cycle

Stage-based provision

Conclusions

1. Collection and use of longitudinal data in accordance with the life course perspective is increasingly recognized as a key tool for science and policy - hence the widening investment in longitudinal research resources across the world.

2. Such multi-disciplinary enquiry enables identification of the key features of changing social, economic, political and environmental contexts impacting on child development and to chart their long term effects.

3. GUI’s potential value will be enhanced by the growing comparative opportunities for identifying the key formative influences on development to which policy can be directed across the life course in Ireland.